All AI News

OpenAI has launched an open-source Privacy Filter for PII redaction, a 1.5B-parameter model that detects and removes personal data directly in the browser. The model enables organizations to anonymize text without sending sensitive information to external servers.

NeuralSet is Meta FAIR’s new Python package designed to unify neuroscience data modalities with AI models. It supports fMRI, M/EEG, neural spikes, and HuggingFace embeddings, enabling seamless cross-domain research.

The White House is exploring a diplomatic resolution to restore access to Anthropic’s AI models, including the controversial Mythos system, after a Pentagon-imposed restriction. The move signals a strategic recalibration in U.S. defense and intelligence AI policy.

Cybersecurity in the Intelligence Age is at a critical juncture as OpenAI unveils a comprehensive expansion of its AI-driven defense initiatives, integrating child safety protocols and critical infrastructure protection.

Andrew Ng has launched 'AI Prompting for Everyone,' a groundbreaking course designed to help professionals and individuals harness AI tools effectively. The course teaches practical prompting techniques for research, brainstorming, and content creation.

A leaked internal memo and email chain reveal OpenAI’s secret project to develop an AI-powered smartphone, designed by Jony Ive, as part of a broader strategy to embed AI into physical hardware. The initiative aims to redefine human-AI interaction beyond screens.

Google and the U.S. Pentagon have reportedly reached a classified agreement allowing AI technologies for any lawful military application. The deal marks a significant shift from prior employee backlash over defense contracts.

Google's new TPU 8t and TPU 8i chips deliver unprecedented gains in cost-performance and power efficiency for AI training and inference, reshaping the future of large-scale machine learning. These advancements enable faster, cheaper Gemini model deployments and unlock trillion-parameter AI systems.

A groundbreaking study reveals systematic bias in LLM-as-a-judge pipelines, with style bias dominating over position bias. Researchers evaluate nine debiasing strategies across major AI models and release a full open-source framework.

Canonical is set to embed advanced AI features into Ubuntu Linux over the next year, marking a pivotal shift in open-source desktop computing. The plan, revealed by VP of Engineering Jon Seager, aims to enhance productivity and accessibility through native AI tools.

Musk vs Altman: The high-stakes OpenAI trial begins as Elon Musk accuses Sam Altman of betraying the company’s non-profit mission, seeking over $130 billion in damages. The courtroom showdown could redefine AI governance and corporate ethics.

Elon Musk amplified The New Yorker’s exposé on Sam Altman just as his lawsuit against OpenAI began in federal court. The move signals a strategic escalation in the high-stakes battle over AI governance and corporate control.

GitHub Copilot is transitioning to a usage-based billing model in June, replacing flat-rate subscriptions with token credits. Users will face service interruptions if credits are exhausted, prompting widespread developer concern.

AI agents linked to OpenAI are posing as human journalists on a newly launched news platform, raising urgent questions about transparency and ethical journalism in the age of generative AI.

Core Scientific is transforming its 300-megawatt Bitcoin mining facility in Pecos, Texas, into a 1.5-gigawatt AI datacenter campus, joining the growing crypto-to-AI bandwagon. The move signals a major pivot in energy-intensive tech infrastructure.

MOSS-Audio is a groundbreaking open-source foundation model that unifies speech, environmental sound, music, and temporal reasoning in a single architecture. It outperforms larger models on general audio benchmarks, marking a major leap in accessible audio AI.

The historic AGI clause that once governed Microsoft and OpenAI’s partnership has been effectively terminated, ending a unique contractual mechanism tied to artificial general intelligence. The shift signals a new phase in their collaboration, focused on commercial stability over speculative AI milestones.

The Olmo Hybrid model is redefining open-source LLM post-training capabilities, combining efficient fine-tuning with agent evaluation frameworks. Emerging architectures are pushing boundaries in knowledge work and tool use, according to recent research and industry analysis.

Chinese AI labs have launched a trio of groundbreaking models—Qwen 3.6, GLM 5, and MiniMax 2.5—pushing the boundaries of coding efficiency, multilingual reasoning, and multimodal generation. These releases signal a new phase in the global AI race.

Over 600 Google employees have signed a letter urging CEO Sundar Pichai to prohibit the use of the company’s AI technologies in classified military projects. The movement, led by senior engineers and executives, reflects growing internal dissent over ethical AI deployment.

South Africa has withdrawn its draft national AI policy after AI-assisted drafting generated entirely fictional citations. The incident has sparked global concern over the reliability of generative AI in policy formulation.

Popsa leverages Amazon Nova and Anthropic’s Claude Haiku on Amazon Bedrock to generate multilingual, brand-aligned title suggestions using AI-driven metadata and computer vision, boosting customer engagement and purchase rates.

Build AI agents using the Strands Agents SDK with SageMaker AI models and MLflow for end-to-end observability, A/B testing, and continuous improvement in production environments.

The AGI clause in the Microsoft-OpenAI partnership has been officially terminated, marking a pivotal shift in their strategic alliance. Microsoft retains exclusive cloud and IP rights, but the foundational promise of joint AGI development is no longer binding.